A Modified Algorithm for DOA Estimation of Coherent Signal with SVD

The research of coherent signal DOA has been a focus of research estimated spatial spectrum estimation. With the singular value decomposition, which presents a modified algorithm (MSVD) to solve signal-related research, i.e., with a new matrix of a certain rule that from the feature vector corresponding to the largest eigenvalue matrix of signals related, then an amendment to this matrix, and finally get some information of signal with SVD. In low SNR conditions, the algorithm can estimate coherence source effectively, it achieves higher estimation accuracy than other methods. Simulation results that MSVD algorithm has better performance than the conventional MSS and SVD algorithms in accuracy and stability; when the SNR is -5 dB and the angular difference of two signal sources is 10°, the successful rate of MSVD algorithm is nearly 100% but those of conventional MSS and SVD algorithms are much lower.